P20.08: Prediction of small‐for‐gestational‐age neonate in a cohort of estimated normal weight fetuses at term

Abstract

Small-for-gestational-age (SGA) neonates are at increased risk of perinatal mortality and morbidity. Fetuses estimated by ultrasound to be appropriate for gestational age (AGA) at term could subsequently become SGA by the time they are born. We aimed to develop a predictive model for the risk of delivering an SGA neonate in a cohort of estimated AGA fetuses at term. A cohort study of singleton fetuses at term with an estimated fetal weight (EFW) between 10th and 70th centile using the INTERGROWTH-21 charts. Major structural anomalies, aneuploidies, genetic syndromes, missing outcomes or stillbirths were excluded. The variables included known risk factors for SGA: maternal characteristics, EFW, fetal Doppler indices and interval to delivery. Logistic regression, ROC curve analysis and Hosmer-Lemeshow tests were used to analyse the data. 8267 AGA term pregnancies were included in the analysis, of which 2023 (24.4%) were SGA at birth. The prediction model (AUC 0.66, 95% CI: 0.65-0.68) included middle cerebral artery MoM (OR: 0.57, 95%CI: 0.50 to 0.66), interval to delivery >2 weeks (OR: 2.04, 95%CI: 1.82 to 2.28), and multiparity (OR: 0.42, 95%CI: 0.38 to 0.47). The EFW at last scan alone had a poor predictive value for SGA neonate (AUC: 0.52, 95% CI: 0.51 to 0.54). The multivariate model also had suboptimal fit according to Hosmer-Lemeshow test (P=0.002), but outperformed EFW alone for the prediction of SGA neonate. Even though the prediction model shows modest precision and poor calibration for assessing the risk of SGA at birth in a cohort of estimated normal weight fetuses at ultrasound, the multivariate model outperforms EFW in the prediction of SGA neonate in a cohort of AGA fetuses at term.